Brought to you by Blue Chip

Blue Chip Cloud

Why Blue Chip Cloud is resilient and flexible

A look at the security and agility that drives Blue Chip Cloud. Phil Lynch, our Architecture and Automation lead, looks at what underpins our outstanding cloud service.

Phil Lynch

Architecture and Automation Lead at Blue Chip

How are we able to provide a resilient and flexible IBM Power Cloud to our customers? Phil Lynch, our Architecture and Automation Lead, speaks on our delivery of cloud-based environments, which takes in the need for high security and rapid deployment.

So today I'm going to be talking about the Blue Chip Cloud and what makes it a flexible and resilient solution.

Our cloud has been designed with a Software-Defined Data Centre at its heart. Taking a software-defined approach gives many benefits over traditional data centre design. One of the key benefits that I'm going to go through first, is around speed of deployment.

Blue Chip has been working for some time on creating a set of standard patterns that significantly reduce the amount of time it takes to design and deploy customer environments. What I've drawn up on the board is the bare bones of 'pattern 1', which is by far our most common pattern.

Within this pattern, we start off by first of all focusing on the IBM Power systems. So for the majority of our customers, they tend to have an IBM i LPAR, so what we would do with our automation software, is start with the IBM i, and we will get it to connect to our storage system and deploy out to the necessary storage on demand. Once that's complete, we then connect the IBM i to the local network, and then this piece of the infrastructure is unavailable.

At the same time, we would have a virtual machine, over this side, which could be running say Windows or potentially Linux. We will use our orchestration tool to connect out and deploy storage on demand and in a similar fashion, connect it to the local network. This then allows the compute layer to become online and be built very very quickly.

Something which not many of our competitors have managed to do, is actually automate the deployment of the IBM i, at the same speed as automating traditional Windows and Linux virtual machines. So the next part of the pattern is to focus on the security. So Blue Chip will deploy a highly available pair of firewalls. These firewalls are next generation and they offer a large variety of different advanced functionality.

Once the firewalls are online, the final task is to connect that to the internet. Once that is established, the whole of pattern 1 is essentially built and we can then quickly enable customers to connect in from remote sites and get them access into their IBM i LPARs and into their Windows machines and Linux machines very rapidly.

This whole approach means that from a project point of view, activities such as migration can happen much much faster than would happen in a traditional data centre. In this next section I'm going to be talking about another big benefit that the Software-Defined Data Centre brings. This time, talking about the firewall architecture and how Blue Chip has designed our Software-Defined Data Centre to provide the highest level of resilience under all circumstances.

On the board I've drawn up two sites. In site one I've got three hosts that represent a virtual compute layer, and similarly, a site two. When Blue Chip deploys virtual firewalls we would deploy them in such a way that it will operate out of the first site and also we will deploy in the second site. These two firewalls establish a cluster and they perform a highly available service.

With our Software-Defined Data Centre, the highly resilient nature of both sites allows for a host to fail. When that failure event happens, the virtual machine at site one will automatically be recovered on another host that is best suited for it. As soon as this recovery takes place, the cluster will be automatically reestablished. This ensures that within either site, at any moment, we always maintain a highly resilient pair of firewalls. This is very different from traditional data centres, where typically, if you have physical firewalls, if you were to lose either of those, then during the time whilst that device is down, your business will be running with a single point of failure. Whereas with Blue Chip's cloud-based solution, we ensure that your business and your SLAs and your RPOs and RTOs remain unaffected by an outage on one of our hosts.

In this next section I'm going to be talking about the design that makes up our virtual data centre. On the board I've drawn up the two sites again, but this time they are logically connected into as one site, but at site one we have a storage array and similarly at site two, the storage systems themselves are replicated and this is essentially a synchronous replication.

Having synchronous replication allows us to have storage at both sites at any given time. When we deploy virtual machines, we can deploy them at site one and at site two. And, at either site, the machines will connect down to the same network. Our software-defined network has the flexibility to allow us to have networks essentially anywhere we need them.

For the virtual machine, it will connect into storage system one, and it will also have a path across to storage system two. Having this virtual compute layer and the virtual storage layer configured in this way allows a huge amount of flexibility.

Essentially, virtual machines can move between sites, and we can do this under our own control or the system will choose where best to place virtual machines based upon current load. Not only does this give us flexibility of where machines sit, but it also allows us and our customers benefit by not needing to perform traditional DR, because, essentially, the disaster recovery is built into the platform by its design.

Where we have a failure scenario, where for example if we lost site one, and all of this infrastructure was unavailable, then what would happen is that this virtual machine would no longer work at this site and it would be automatically recovered over at site number two. This recovery event would be automatic, and compared to traditional data centre designs, where, typically engineers need to get involved when there's a major disaster, Blue Chip's cloud design allows the automated recovery of those machines.

In this final section I'm going to be talking about hybrid cloud. This is something that more and more of our customers are looking to adopt. On the diagram here, I've drawn up this site which represents Blue Chip. Within Blue Chip, we have the similar pattern one deployment that I drew up earlier.

On the right hand side - represents a public cloud provider, and within there, a virtual environment that the customer has machines running within. What we have in the middle here, is essentially an interconnect that allows Blue Chip's data centre to connect out to a variety of different public clouds. Essentially by connecting into the interconnect and then down, into this environment here, that establishes connectivity and it allows our operations centre to have visibility of the network that is here in this space here, and similarly, networks down here.

This is a visual representation of this network from the public provider. This view that we get is an enhanced view where we can see status of machines, both within Blue Chip data centres and also out in the public cloud. This allows us to provide many services through a single contract. Our support teams can do this no matter where the machines are.